594 MURID.E— EPIMYS 



of Hossack, and was carried out from a biological and plague point 

 of view ; he likewise criticises the systematists, but in his paper many 

 facts supporting the view that the "varieties" of rattus breed true 

 in the country districts will be found ; thus rats in large numbers were 

 collected from sixty -nine villages in the Punjab, but white - bellied 

 specimens were only found in three rather widely separated localities, 

 two of them in the Amritsar district, and one in the Lahore district. 



The Egyptian races of rattus have also been studied by Bonhote 

 {op. cit., 1910), who finds only two forms evident there, viz. : — 



ii) E. r.frugivorus} Rafinesque, with white under parts and light- 

 coloured feet, the hind foot being usually 35 mm. long. 



(2) E. r. alexandrinus, Geoffroy, with slate-grey under parts and 

 dark feet, the hind foot being usually 33 mm. long. 



Owing to the propinquity of forms, these races are in Egypt very 

 much mixed up together, and in practice merely represent extremes of 

 a varying series. Bonhote regards these two forms as the analogues 

 of the Indian rufescens and griseiventer sub-groups ; he finds by 

 measurements that a third type, indistinguishable by colour, but with 

 a hind foot of 29-30 mm., exists, and he suggests that this corresponds 

 to the Indian jalorensis sub-group, but in Egypt no difference of habit 

 is visible. 



A recent examination of the material in the British Museum from 

 all parts of the world other than the Oriental region, amounting to 

 more than 200 specimens, showed that, adopting Bonhote's definitions, 

 frugivorus could always be distinguished from alexandrinus ; but some 

 difficulty was found in separating the darker specimens of alexandrinus 

 from the paler phases of r. rattus. Bonhote (1910) showed that when 

 the hind foot measurements of a large number of individuals were 

 plotted as curves, either lor frugivorus or alexandrinus separately or for 

 both races together, the curve in each case showed three apices, viz., at 

 35, 33, and 30 mm. In frugivorus the major apex for both sexes 

 occurred at 35 ; in alexandrinus it occurred at 33 for males and at 35 

 for females ; the minor apex at 30 was regarded as an indication of the 

 latent presence of a representative of the Ynidix^xi jalorensis group. The 

 material in the British Museum similarly treated has given a similar 

 result, save that for both sexes of alexandrinus the major apex occurs 

 at 33 mm. Such facts appear to indicate, according to Bonhote, that 

 the three forms are in each case mutations, indestructible and ready 

 to develop whenever a suitable , environment offers. The curve for 

 r. rattus is, however, much more complex, since it shows for both sexes 

 large apices at 34 and 36 and a smaller one at 38 ; for males minor 

 apices occur at 30 and 32 mm. 



' Bonhote uses Savi's name tectorum for this form ; " tectorum " is, however, ante- 

 dated by Musculus frugivorus, Rafinesque. 



